Talk:Profitability/@comment-217.162.151.111-20110716145518/@comment-217.162.151.111-20110718184233
Hi, I improved the simulator and plotted some data. You can get it here: http://www.mediafire.com/?n56behiie75hyai The description is below, but first I want to comment on your statements. Reaction to comments CMONYALL: > "We concluded that there is no way we can possibly take this frequency into account for every structure. Actually even your program cannot do this." It does now. I agree, that the collection frequency decreases over time, but the ranking does not change if the duration between collections is above the maximum accumulation duration of a building. > "The result is a superb overview of the structures' relative profitability. Nothing to do with a 'definitive guideline' as you call it. Just hard facts." Which is great, I did that as well, but I realized that the frequency has an extreme impact. Since I don't wanna change my collection habits, I investigated which building has a high and _stable_ profit. As you will see in my graphs, it actually is a big difference. Concerning comments: I developed the tool for my own use. Since I discovered that the collection frequency has a big impact, i shared it with the world. ;-) The new version has some explanation. If you still feel it to be insufficient, you're free to document the code... ;-) Concerning probability: I used my current ratio, since I developed the tool for myself. But it was saved in a variable to easily change it. Concerning the Yellow Buck: I always collect it. I don't know how anyone else handles this issue. You can't do anything with the building if you don't click on that buck, so what's the reason not to click on it? The graphs are very stable with the yellow buck. The rule is the same for every building as well: collect it when the buck appears. > "So, long story short: you complain about our table and charts, but actually, your simulator is even worse. " I improved my simulator and I trust the output. If you don't like the experimental approach, just say it honestly and politely or suggest an improvement. Despite on this page are some interesting facts and figures, (which were hardly collected, so thanks a lot for the responsible persons!) considering the influence of the collection frequency, these facts need to be processed and my approach seems more suitable. BTW, I intentionally omit the nice sentence "I hope you like it", because it's cool if you like it and I'm delighted then, but I don't care if you don't, but I'm always open for any serious suggestion. Pestilence Rex: Thanks for your support! You don't have to run it anymore, if you agree to my parameters. You can just look at the graphs. ;-) Description & Details All Graphs: Collect on Yellow Bucks, repair probability = 1/14. http://img708.imageshack.us/img708/7231/simulationp.png Simulation: Raw data. 10000 iterations of money collection for every minute +/- 30 on the X-axis. Example: x = 240 minutes means y = average dollar per hour for 10000 money collections, each randomly after between 210 and 270 minutes. http://img638.imageshack.us/img638/7236/simulationfilter.png Simulation_Filter: Filtered graph, see spreadsheet for details. More graphs in the 7zip-file. The income seems always pretty stable after 10000 iterations.